The goal of this research is to develop improved assessment protocols that afford a quantitative and analytic evaluation of speech impairment in children and adults with neurological disorders. Speech impairments (dysarthrias) will be studied in persons with amyotrophic lateral sclerosis, Parkinson's disease, stroke, cerebellar degeneration, cerebral palsy, and developmental speech disorders. Improved evaluation of speech intelligibility is a particular focus of this work, but issues of speech and voice quality also are addressed. The methods to be used are a combination of standard clinical assessments (such as rating scales), intelligibility evaluations, and computer-based acoustic analyses. Specifically, the methods include: perceptual ratings of speech by experienced clinicians, quantitative assessment of intelligibility, a multiple-parameter acoustic analysis, computer correction of speech abnormalities through LPC resynthesis of the acoustic signal, and derivation of vocal tract shape from acoustic parameters. Work in all these areas will be based on recordings of speech samples from a large number of individuals with dysarthria. One product of the research will be a library of clinical profiles including intelligibility scores, phonetic feature analyses, ratings of speech/voice quality, acoustic measures, and neurological diagnosis. Particular attention will be given to the influences of subject age and sex on the characteristics of dysarthria for a given neurological diagnosis. The assessment protocols will be implemented on microcomputers and designed to be incorporated in clinical practice. The research also will contribute to the development of expert systems for the rating and classification of dysarthria.